Data-driven versus data-informed – the data scientists’ perspective
When we first got access to huge amounts of data and incredible processing capabilities, the whole world seemed to have swung towards data. Becoming the Holy Grail of management and strategic decisions, data gained unprecedented primacy in how executives approached decision making. For years, it was en vogue to be data-driven: making decisions and developing strategies based on the analysis and interpretation of data, also known as hard, unbiased facts. Until it turned out that this wasn’t working.
In my research, I speak to many data scientists, people who make data and run analyses that are used in decisions and strategy-making. The one thing I keep hearing over and over again is that they themselves aren’t data-driven. If a data scientist thinks that data is the oracle, and blindly follows what the results of number-crunching indicate without any critical thought behind it, well, that’s not a good data scientist.
Good data scientists and effective data science teams are data-informed, that is they take into consideration data, research, experience, and personal insights to select, analyze and present data. Some of those I spoke with emphasized that part of their job is to make sure that the decision-makers who then consume the results of their work know not to treat it as gospel, but rather be similarly data-informed. Take the work that the data science team produced, add your own experience and critical thinking, and then use the data – this is how analytics products should be used.
The problem is that managers and executives aren’t really trained in how to consume data science products. All too often they believe that data is the answer, and they fall into being data-driven in their decisions. We know this is not the best approach, and we know it straight from data scientists themselves. But what we lack is solid education on how to use data and analytics outputs in business contexts, and this includes teaching managers how to draw from data, experience and critical thinking. Perhaps for too long we’ve let the future decision-makers believe that data holds all the answers, and it’s time to show them how to be data-informed?